Views please on £280k investment portfolio

aroominyork wrote: »

The way I calculated historic return was to...

... was to look at a relatively short period in which:

a) India and Japan surged with new governments shaking things up by implementing equity-supportive policies under Modi and "Abenomics", including taking interest rates negative in Japan for a sustained period which hasn't yet ended;

b) major world economies recovered from being on the edge of recession post- credit crunch and the market drops of 2011;

c) all international stocks and bonds including the UK FTSE stocks which have high overseas revenues got a huge boost in sterling terms as the pond weakened significantly against a basket of international currencies especially USD;

d) the lowest interest rates witnessed in recorded history were maintained in the UK, US, Europe, Japan and other developed markets (and in several cases lowered further), boosting equity valuations through cheap borrowing and the enduring lowered yields on bonds;

e) investment grade bonds (and other riskier bonds) delivered total returns well in excess of their actual cash yields, being propelled upwards for the tail end of a pretty much one-way three-decade bull market;

f) UK consumers, discouraged from saving, increased their debt relative to incomes in the face of minimal real wages or productivity growth so corporate earnings got a free boost and property prices pushed to increasingly unsustainable levels; inflation is only now filtering through and the short to medium term damage from the Brexit effects of corporate uncertainty on jobs, profit and investment have yet to be fully felt;

g) the US became self sufficient in energy at crazy low prices after a one off boost from discovering how to take advantage of shale and fracking; Trump got voted in on a promise of abolishing taxes and eliminating regulations in all major industries, but his hundreds of millions of wealthy consumers haven't had it proven to them that he is running on ego alone and the trillions of dollars has to come from somewhere; interest rates are ticking up only delicately so far and QE is yet to be unwound.

So, you'd expect a global portfolio of stocks and bonds to do pretty well over five years in that environment. But given 10-12% after fees annualised is very high as a long term objective for a portfolio of only 60-70% equities, it would seem laughably optimistic to have the next five years deliver that, from where we are now.

I have just run a calculation of risk by multiplying each fund’s FE by its allocation (eg for Ballie Gifford FE 77 x 0.1 = 7.7); the total shows a weighted FE across the portfolio of 75.4. It also revealed that 38.7 of FE risk – ie over half the total risk – lay in three funds (Royal London, Japan, India) which between them hold 35% of the portfolio. That is too much risk in a small number of funds.

Obviously all these calculations have many caveats but I’d be interested to hear views on whether the FE methodology is a sound way of calculating risk across the portfolio? For example, the FE looks too high – I would want to bring it to about 70 – and has too much risk concentrated in a small number of funds.

No, it's not a sound way of looking at things.

The FE risk score is a pretty basic measure. It uses FTSE100 as the "100" score and gives a volatility score relative to that, using data only from recent years (3 or so?) and within the data being compared, weights recent data more heavily than older data. There is some logic in its approach, especially when trying to assign a score to as many of the available funds as possible, which will include funds that have not been running many years yet. But it can give spurious results over different time periods - especially given FTSE is weighted heavily to certain industry sectors and they are not looking long long term anyway, in their quest to use recent and "relevant" data. The most recent data is not always what is"relevant" when looking at the potential of a fund.

I can well believe that both the Japan smaller companies vehicle you selected from Baillie Gifford, and the India fund, should have higher risk scores than FTSE100. They are both investing in the stock exchange of a single geographic region, and one is smallish companies while the other is an emerging market. FTSE100 is single-stockmarket too, and limited in coverage of industries, with currency risk, but it's components are generally multinational giants.

So, you'd expect them to have more risk than FTSE 100 and bonds to have less risk but to just multiply them out by the percentages held is to presume they are accurate and useful measures (rather than interesting but inherently flawed measures) in the first place.

Also just grabbing the scores which were calculated in isolation and then adding them, gives no sense of interactions and correlation between the fund components

Say for example India and Japan were great rivals and their stock markets were negatively correlated. So if Japan equities were to rise by "risk free cash rate plus x%", India would fall, to "risk free cash rate less x%", and vice versa. For your combined portfolio of 50:50, which you would rebalance every year to bring them back to 50:50,, they would perfectly cancel each other out whoever was on top for that period, and your net return of the gains and losses would just be cash return, plus X gained on Japan minus X lost on India... equals cash return. The most smooth and stable return you could get.

Whereas in your analysis you just say ooh they are both 1.5 times as volatile as FTSE, so the riskiness of this part of my portfolio is 150% of the FTSE, I could expect much steeper dips but hopefully in the long term much higher returns. At least that's what your risk score analysis told you, when in fact they were a perfect complement and cancelled each other out.

It is same with equities vs certain types of bonds. You hope the bonds go up in an equities crash and the equities go up when bonds are not delivering much (or delivering negatives). That negative correlation helps your portfolio. If you just add up and mash together the fact that the two things can be volatile, you will probably think that the whole thing will be very volatile, when actually having two volatile assets move in different directions at different times is the way you take advantage of the diversification inherent in the portfolio and sell one high to top up the other one that's low. Smoothing overall returns and reducing the overall standard deviation from the mean rather than doubling it up.